Grand average of Time-Frequency map (MEG-based sources)

Discussions
,
1

Hello,

I've referred to this tutorial (https://neuroimage.usc.edu/brainstorm/Tutorials/VisualGroup#Group_analysis:_MEG.2FEEG)
and I'm wondering that whether I can use the following method to calculate the grand average of TF map of MEG data across subjects (I'm using scouts from atlas):

  1. average sources between runs (by trial group [subject average])
  2. standardize these sources (baseline normalization)
  3. extract the absolute values
  4. project on default anatomy
  5. spatial smoothing
    then there will be the final processed sources for each subject
  6. using these processed sources >process source > frequency > time-frequency (morel wavelets) (use scouts)
    then there will be the TF maps for each subject
  7. using these TF maps do average across subjects (by trial group (grand average))

Does this method work?

Looking for your help!

2

This is not right. TF maps should not be computed:

  • On trial averages. Averaging will remove some of the induced response: the brain activity in higher frequencies that is not strictly time-locked to the stimulus, and not aligned in phase across trials.
  • On sources that were already rectified (your step 3)

  1. Within each subject, compute a TF map (using Scouts) for all the files trial group from all Runs. The result will be the Subject average TF map. As shown in here. You will need to specify if the Scout function is applied before or after the TF transformation).
    At this point the TF maps are in the same space (Scout space)

  2. Standardize the Subject average TF maps

  3. Compute the grand average with the Subject normalized TF maps

3

Got it! Thank you for your reply!

4

@Raymundo.Cassani Sorry to disturb you again. I'm also wondering that, if I want to compute (1) the connectivity of each subject, (2) the grand average of connectivity (3) and statistic analysis (against baseline), should I use the projected source in the "Group analysis" file at the top?

5

As with TF, connectivity metrics on source space should be not computed on rectified sources (which is the case with sources that have been projected on the Default anatomy)

6

@Raymundo.Cassani Thank you for answering! There is another point that I want to make sure:

Does it also means that, if I use atlas scouts to calculate connectivity, then I don't need to project the connectivity result on template? Just as the TF map calculated using scouts?

7

That is right, as the same scout closely corresponds to the same brain area for different subjects.

8

Thank you very much!